An autonomous path-planning strategy based on Skinner operant conditioning principle and reinforcement learning principle is developed in this paper. The core strategies are the use of tendency cell and cognitive learning cell, which simulate bionic orientation and asymptotic learning ability. Cognitive learning cell is designed on the base of Boltzmann machine and improved Q-Learning algorithm, which executes operant action learning function to approximate the operative part of robot system. The tendency cell adjusts network weights by the use of information entropy to evaluate the function of operate action. The results of the simulation experiment in mobile robot showed that the designed autonomous path-planning strategy lets the robot realize autonomous navigation path planning. The robot learns to select autonomously according to the bionic orientate action and have fast convergence rate and higher adaptability.
Based on the simplified fish motion model, a robot fish which could detect the oil leakage point of pipeline was designed by the method of single-joint driving. The Hawkeye OV7725 was used to design the image acquisition module to obtain the current movement of the fish and the current pipeline situation and the collected data was processed for making the relevant decisions to achieve the direction of movement control with the STM32 microcontroller. On the basis of binarization image centroid method, the image recognition algorithm was studied. By using the coordinates of the white point in the two-dimensional array, a linear regression equation which can reflect the distribution trend of the white point in a frame image was designed and the motion direction of the current robot could be detected. Since the linear regression equation converge to the characteristics of discrete data points, the oil leakage point inside the white area of the image could be detected. Experiment results showed that the robot fish can effectively complete the oil spill point detection task.